Performance paper - Enanpad 2011


Published on

Published in: Education, Technology
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Performance paper - Enanpad 2011

  1. 1. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia OrganizacionalUNCTAD’s Degree of Internationalization and Its Effect on Subjective andObjective Performance: Evidences from Brazilian TNCsLívia Barakat, Sherban Leonardo Cretoiu and Jase RamseySummaryBrazilian Transnational Corporations (BTNCs) have been increasingly engaging inforeign direct investment over the past decade. With that in mind, there is a wide rangeof strategic choices that BTNCs must undertake when internationalizing in terms ofentry modes, degree of ownership, control, centralization, among others. That results indifferent levels of internationalization. For years, several scholars have attempted toestablish a reliable measure of the degree of internationalization. Whereas some rely ona single-item usually based on foreign sales, others propose a multidimensional measurethat considers FDI, employment and geographic dispersion, among other indicators.Therefore, this paper aims to assess BTNCs transnationality index based on twodifferent measures: the UNCTAD’s transnationality index and a composite index thatincludes other measures such as number of countries and international experience. TheUNCTAD’s index considers the average of three dimensions: foreign assets divided bytotal assets, foreign revenues divided by total revenues and foreign employees dividedby total employees. We also assess the impact of the degree of internationalization onboth objective and subjective performance. In order to test our hypothesis, we propose amodel of internationalization assuming a positive impact of the degree ofinternationalization on foreign performance. An empirical study was conducted with 41BTNCs from a population of 71, regarding its international activities in 2008 and 2009.Data was entered and analyzed in the softwares SPSS and Amos using StructuralEquation Modeling and Tests of Multivariate Assumptions. Our results show that themore internationalized a firm is, the better it performs overseas. Firms with a higherdegree of internationalization were found to be more satisfied with foreign sales, salesgrowth, profits and market share. Additionally, internationalization leads to a higherpercentage of foreign profits over total, and an increase in foreign return on sales. Onthe other hand, return on assets is not significantly impacted by the transnationalityindex. Furthermore, we showed that the UNCTAD’s transnationality index is morereliable in this context than a construct that also includes other internationalizationmeasures. Thus, firms may see the internationalization strategy as a way to enhanceforeign performance. Finally, internationalization increases objective performance andalso executive’s satisfaction with performance. Yet, the effect of internationalization isstronger on objective measures.Keywords: Internationalization, Subsidiary Performance, Degree ofinternationalization. 1
  2. 2. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia OrganizacionalINTRODUCTION Internationalization has long been discussed in the strategic managementliterature (eg. Dunning, 2000; Johanson and Vahlne, 1977), as a way of diversifyingbusiness and creating value. Arguably, transnational companies (TNCs) pursue superiorperformance when engaging in foreign direct investment (FDI) (eg. Sharma, 1998).However, TNCs have different strategies in terms of entry mode (eg. Kogut and Singh,1988); location (eg. Goerzen and Beamish, 2003), centralization (eg. Davidson, 1984),and ownership (eg. Hennart and Reddy, 1997). This wide range of strategic choices thatTNCs must undertake when internationalizing result in different levels ofinternationalization. With that in mind, several scholars have attempted to establish areliable measure of the degree of internationalization. Whereas some rely on a single-item (eg. Geringer, Beamish and da Costa, 1989), usually based on foreign sales, otherspropose a multidimensional measure (eg. Ietto-Gillies, 1998; Sullivan, 1994; UNCTAD,1995) that considers FDI, employment and geographic dispersion, among otherindicators. Although the choice of an ideal measure seems to depend upon the contextstudied (Ietto-Gillies, 1998), it is important to assess the internationalization level offirms in order to understand the patterns and effects of global strategies. In Brazil, forinstance, TNCs have been increasing FDI 1 over the past decade and performance isclaimed to increase accordingly (Loncan and Nique, 2010). Thus, this paper addressesthe following questions: do more internationalized TNCs perform better than lessinternationalized? What is a good measure of the degree of internationalization in theBrazilian context? Does the degree of internationalization impact differently objectiveand subjective performance? Therefore, this study aims to compare two differentmeasures of internationalization: the UNCTAD’s transnationality index and a compositeindex that includes other measures such as number of countries and internationalexperience. Moreover, we propose and test a model that assumes internationalization topositively impact both foreign objective and subjective performance of firms.THEORETICAL BACKGROUNDDegree of internationalization In the last decades, many studies have been developed to find an ideal way tomeasure firm’s internationalization. The first attempts to quantify internationalizationuse a single-item approach which usually considers firms’ foreign sales (eg. Collins,1990; Dunning, 1985; Geringer, Beamish and da Costa, 1989; Grant, Jammine &Thomas, 1988). Others assess internationalization by the ratio of foreign assets over total(Ramaswamy, 1993). Hence, the majority of the literature on the degree ofinternationalization is based on a dichotomy of foreign x total (eg. Davidson, 1984). Nevertheless, over the past years new approaches to the degree ofinternationalization have been developed. As a result, different patterns ofinternationalization, not limited to the financial dimension and not only based on asingle criterion (e.g. foreign sales), have gained attention. Lu and Beamish (2004), forinstance, assess internationalization by two variables: a firm’s number of overseassubsidiaries and the number of countries in which a firm had overseas subsidiaries in agiven year.1 Balance of Payments. Central Bank of Brazil. 2
  3. 3. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia Organizacional According to Sullivan (1994) internationalization is a multidimensionalphenomenon which has the following attributes: performance (Vernon, 1971), structural(Wells, 1972) and attitudinal (Perlmutter, 1969). Thus, in order to increase reliability ofIB studies, multi-items’ measures started to be used as means to control themeasurement errors and to cover the different aspects of the process (Dorrenbacher,2000). Fisch and Oesterle (2003), for instance, propose a broader approach in terms offirms’ geographical spread of activities cultural diversity. Among the emergent studies using composite indicators, we outline three: (1)The DOI (Degree of Internationalization) developed by Sullivan (1994); (2) TheUNCTAD’s three-dimensional index (1995); (3) and Ietto-Gillies’ TransnationalActivity Spread Index (1998). The Degree of Internationalization developed by Sullivan (1994) consists on alinear combination of variables that covers three aspects of internationalization:attitudinal, performance and structural dimensions. The DOI index is composed by thefollowing indicators: the ratio of foreign sales to total sales, foreign assets to totalassets, foreign subsidiaries to total subsidiaries, international experience of topmanagers and physic dispersion of the international operations. For instance Cadogan,Kuivalainen, and Sundqvist (2009) base on Sullivan’s geographic diversity dimensionto compose their exporters’ degree of internationalization. Bobillo, Lopez-Iturriaga, andTejerina-Gaite (2010) also use elements of the DOI in a study of firm performance andinternational diversification. The three-dimensional index developed by UNCTAD was launched in theWorld Investment Report (1995). It combines three ratios: foreign sales to total sales,foreign assets to total assets, and foreign employment to total employment. Thetransnationality index averages the tree dimensions in order to balance different types ofinternationalization among various industries. Tuselmanna, Allenb, Barretta, McDonald(2008) for instance, have employed it to assess the importance of industryinternationalization in shaping the strength and nature of the country-of-origin influencein employee relations approaches of US subsidiaries. Although still little explored byscholars, several studies have also considered it to draw an internationalizationframework or to build its own measure (eg. Outreville, 2008; Ruzzier, Antoncic andHisrich, 2007). Furthermore, the UNCTAD’s transnationality index has been adoptedby business schools in several countries to study internationalization of nationalcompanies. Finally, Ietto-Gillies (1998) developed the Transnational Activity Spread Index,which combines elements of UNCTAD’s index with the physic dispersion of theforeign activities, as developed by Sullivan (1994). Ietto-Gillies’ approach considerstwo dimensions (i) the intensity level of internationalization in relation to the overallsize of the activity or economy; and (ii) the level of geographical extensity of theinternational activities. This index has been used for instance, as way to assess therelationship between multinationality and innovation (Frenz, Girardone and Ietto-Gillies, 2005). Yet, regarding internationalization measurement, the author (p.17-18)states: “There is no single way of assessing the degree to which companies, industriesor countries are internationalized: it all depends on what patterns and aspects ofinternationalization we choose to emphasize […]”. Therefore, the measurement choice in this study takes into account the context ofrelatively young Brazilian multinationals and the inherent difficulties in collecting datawith international managers. Thus, the UNCTAD’s index is the most feasible of the 3
  4. 4. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia Organizacionalthree approaches, due to its simplicity – the index considers only three dimensions:assets, revenues and employees. The UNCTAD’s transnationality index has been employed in studies ofBrazilian Transnational Corporations (TNCs) by Fundação Dom Cabral since 2006 (eg.Ramsey, Barakat, Cruz and Cretoiu, 2010; Ramsey and Barakat, 2009). The researchproject Transnationality of Brazilian Companies Ranking annually assesses the degreeof internationalization of Brazilian-owned firms in order to monitor BTNCs’ foreignactivities and to test hypothesis regarding IB theory. Thus, among several topics in IB research, some questions have particularlycaught attention of this study: Does the internationalization increases BTNC’sperformance? If so, to what extent? Is the UNCTAD’s transnationality index effectivein measuring the degree of internationalization of Brazilian TNCs? Due to the growing involvement of Latin American firms in the global scenario,international business has become a topic of interest in the international businessliterature in Brazil. For instance, Loncan and Nique (2010) showed that the moreinternationalized the company (in terms of foreign sales over total sales) the better itsperformance (returns on assets). However, the author only studied five companies andused a single indicator for each concept. Thus, this study aims to further explore thisrelationship by surveying a larger sample, testing multidimensional constructs andadding perceptual measures of performance.Internationalization and Firm Performance Many studies have discussed the effects of internationalization on firm’sperformance. In fact, through global strategic planning, risk management and focus onunique advantages, MNEs have been increasing return on foreign investment (Dymsza,1984). The importance of studying performance of subsidiaries lies in the fact thatinternational firms develop firm specific advantages that may lead to superior returns(Bouquet, Morrison and Birkinshaw, 2009). Therefore, MNEs’ strategic choices aredetrimental to success. Moreover, Sharma (1998) shows that decisions on entry modeaffect post-entry performance of multinationals, which is also influenced by factors suchas industry attributes, advertising intensity, and relatedness with the parent firm. Other studies show that performance is a function of location decisions. Goerzenand Beamish (2003), for instance, argue that the larger the portfolio of internationaloperations, in terms of asset dispersion, the higher MNEs performance. Accordingly,when cultural distance increases, international joint ventures (IJVs) have longerdurations and are less likely to end (Park and Ungson, 1997). Despite little emphasis from scholars, the effect of the degreeinternationalization on performance has also been studied. Lu and Beamish (2004)showed that internationalization moderates the relationship between geographicdiversification and firm performance. The relationship found was non linear, in a waythat at high and low levels of internationalization, the extent of geographicdiversification was negatively associated with firm performance. Conversely, atmoderate levels of internationalization, performance increases as geographic diversityincreases. As several studies in the literature state, internationalization strategies oftenleads to superior performance (eg. Buch, Kleinert, Lipponer, and Toubal, 2005). 4
  5. 5. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia OrganizacionalConsidering that most of international strategies require some level of commitment toforeign markets, we hypothesize that: H1: The more internationalized the MNE, the better its foreign performance. However, it is important to point out that two firms with similar objectiveperformance may evaluate its success differently. Managers may be more or lesssatisfied with performance depending on how it compares to expectations (Oliver,1998) or competitors (Shoham, 1998). For that reason we assess both objective andsubjective performance.Objective measures of performance Objective indicators are represented as numbers or percentages directly found infinancial statements, balance sheets or market/sales reports. Recent studies have beenusing a combination of multi-items (eg. Andersson, Forsgren and Pedersen, 2001). Leland Miller (2008), for instance, employ measures such as stock price, stock returns, andearnings before interest and taxes (EBIT) to assess performance of international cross-listed companies. Taggart and Taggart (1999) also use sales based measures such asmarket share and exports growth to show the linkage between exchange-rate stabilityand performance, as a dimension of firm competitiveness. Other authors such as Miller and Eden (2006) use return on assets (ROA), that isthe ratio of net income to total assets, whereas Makino and Isobe (2003) employ returnon sales, which is the ratio of net income to total sales. Bouquet, Morrison andBirkinshaw (2009) also use return on equity. The authors show that the more executivesinvest time and effort in activities, communications, and discussions to understand theglobal marketplace, the better the firm performance. Thus, each context requires adifferent set of the various measures of performance in the literature. Therefore, our first subset of hypothesis is: H1a: The more internationalized the MNE, the better its foreign objectiveperformance.Subjective measures of performance Despite numerous ways to objectively measure performance, difficulties inobtaining and dealing with companies’ numbers still remain in IB research. First,financial statements are usually confidential and restricted to internal control(Woodcock, Beamish, and Makino, 1994). Lack of objective information is thusresponsible for reducing response rate in IB performance empirical work (Wall, Michie,Patterson, Wood, Sheehan, Clegg and West, 2004). Furthermore, objective measures are not easily comparable, once companiesfrom different industries and sizes may have discrepant values, but not necessarilyoutperform each other. This argument is especially relevant in emerging markets, whereinternational profit, growth and market share may be meaningless in the short run(Pangarkar, Klein, 2004). Because MNEs have distinguished reasons to entering inforeign markets, its performance should be measured taking into account firm’sobjectives and expectations. For that reason, Anderson, Forsgren and Pedersen (2001)employ perceptual questions to assess both market and organizational performance, as 5
  6. 6. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia Organizacionalimpacted by technology embeddedness. With that in mind, managers are moreencouraged to evaluate performance in a Likert scale than reporting confidentialinformation (Sousa, 2004). Geringer and Louis (1991) for instance provided empirical evidence thatsubjective measures are adequate to assess firms’ performance. The findings show thatobjective measures are positively correlated with satisfaction with IJV performance andwith perceptions of the performed relative to initial objectives. Thus, scholars have been increasingly applying subjective measures as acomplement to objective measures and also as an alternative to recurrent issues withobjective indicators (eg. Brouthers, Brouthers and Werner, 2008; Al-Khalifa andPeterson, 2004; Nielsen, 2007). Hence, combining both objective and subjectivemeasures seems a reasonable choice, as previously used in the literature (eg. Cicic,Patterson and Shoham, 2002; Colton, Roth and Bearden, 2010). Thus, our second subset of hypothesis is: H1b: The more internationalized the MNE, the better its foreign subjectiveperformanceMETHODS In this section we present the methods and procedures used to build thequestionnaire, collect data and validate the scales.Data Collection A set of 71 Brazilian groups that have entered foreign markets via FDI werecontacted to participate in the survey. The potential response pool included publicallytraded companies listed or not on the Bovespa (São Paulo Stock Exchange) and privatelimited companies (Ltda.). International managers were asked to fill out a 3-pagequestionnaire regarding their international activities in 2008 and 2009. Thus, 44companies replied, being 41 valid (57.7% response rate). The three not validquestionnaires belonged to either only exporters or companies that could not provide thefinancial data needed. After receiving the questionnaires, the data was verified in publicsources to warrant its authenticity. Note that the study considered groups instead ofindividual business since decision making is centralized in the holding company, whichdeliberates international strategies to all controlled units. Therefore, the informationregards to the groups’ consolidated numbers and locations.Sample Profile From the 41 groups that participated in the study, 90% were private-owned asopposed to state-owned. Respondent firms belonged to various industries:manufacturing (51%), services (44%) and natural resources (5%). Additionally,companies are relatively young in foreign markets: 20% opened the first internationalsubsidiary before 1980, 10% started between 1981 and 1990, 29% started between 1991and 2000, and 32% made the first FDI after 2001. Information from two years of 41groups accounted for 82 data points in the final dataset analyzed.Measurement 6
  7. 7. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia Organizacional The proposed model has three constructs: degree of internationalization as thepredictor of both foreign objective performance and subjective performance. To measure degree of internationalization, we applied the UNCTADmethodology, which considers three indexes: foreign sales over total sales, foreignassets over total assets and foreign employees over total employees. The indexesachieved good reliability, with 0.77 Cronbach’s Alpha. As discussed above, using amultidimensional index balances the different ways of internationalizing since we havegroups from different industries. In general, companies from the services sector have alarge number of employees abroad but relatively low amount of assets. On the otherhand, companies from the manufacturing sector can aggregate high revenues abroad,without necessarily having an enormous number of employees. Furthermore, inspiredby other studies of the degree of internationalization (eg. Ietto-Gillies, 1998; Sullivan,1994) alternative items such as number of countries and international experience wereadded to the construct, in order to compare models. The first item was measured by thenumber of countries (Sundaram and Black, 1992) that firms had FDI in 2009. From anorganizational learning perspective (eg. Hennart & Reddy, 1997) internationalexperience was measured by the number of years since the first international subsidiaryhas been established. To measure foreign objective performance, we used three indicators. The firstwas Ebtida index, which is the proportion of foreign Ebtida as compared to total Ebtida.The second measure was foreign return on sales (ROS), which is calculated by the ratioof foreign profit (Ebtida) over foreign sales. ROS is commonly employed to assessfirm’s operational efficiency and has been applied in international business studies(Geringer, Beamish & da Costa, 1989; Daniels & Bracker, 1989). The third measurewas foreign return on assets (ROA), which is calculated by the ratio of foreign profit(Ebtida) over assets. This indicator has also been used in the literature (eg. Rugman,Yipw and Jayaratnez 2008; Loncan and Nique, 2010) as a measure of investmentefficiency. Using indexes instead of the absolute numbers allows us to assess relativeforeign performance and compare companies from different industries and sizes. Theconstruct however showed low reliability, with Cronbach’s alpha of 0.54. Nevertheless,we decided to proceed with the tests in order to keep a minimum of three indicators perconstruct (Kline, 2005). To measure subjective performance, three indicators were used based on thestudies of Al-Khalifa and Peterson (2004). According to this approach, firms assessperformance based on four elements: sales, sales growth, profit and market share. Thus,firms were asked to rate its satisfaction with each of these measures of performance on afive-point Likert scale. The dimensions were proved to be unidimensional by factorialanalysis and reliable (Cronbach’s alphas was 0.78).RESULTS In order to test the proposed model, we used Structural Equation Modeling in thesoftware Amos. This procedure involves simultaneously testing relationships betweenone or more independent variable and one or more dependent variable. Thus, themethod combines exploratory factor analysis with multiple regression analysis(Tabachnick and Fidel, 2001). The first step was to verify convergent validity. This procedure consisted intesting the significance of the items’ factor loadings (confirmatory factorial analysis) ina model that assumes constructs to covary and not to causally affect each other (Kline, 7
  8. 8. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia Organizacional2005). Subsequently, discriminant validity was achieved once correlations among pairsof constructs were less than unity, and correlations among items were larger thancorrelations among traits (Bagozzi, 1993). Finally, nomological validity was verified bytesting the predicting power of exogenous constructs on endogenous constructs. Thisprocedure involves fixing variances at unity in order to assess the path coefficients(Anderson and Gerbing, 1988). Furthermore, we discuss models’ fit and compare twodifferent models (Balla and Donald, 1988).Descriptive analysis This section presents descriptive statistics and correlation analysis of thevariables in this study. The following table shows means, standard deviations andbivariate correlations between the moderator, independent and dependent variables.------------------------------------------------Insert Table 1 about here------------------------------------------------Structural Equation Modeling Two models were tested in order to compare scales used and to find the bestfitting structure. The first model used the three indexes of UNCTAD (ratios of revenues,assets and employees) as the internationalization measure. Each of the performanceconstructs had three items as presented in the methods section. Results of model 1 areshown in the next figure:------------------------------------------------Insert Figure 1 about here------------------------------------------------ Convergent validity was achieved for most of the items in model 1. The threeUNCTAD indexes are significant (0.1% level) and have high factor loadings, all above0.80. Additionally, all items of subjective performance are significant, though the salesbased (sales and sales growth) reflect subjective performance more precisely thanmarket share and profits. Moreover, from the three items of objective performance, onlyEbtida index is significant at 1% level. Foreign ROS slightly reflects objectiveperformance and is significant only at 5% level. Foreign ROA is not significant at all. Furthermore, transnationality index was found to have a stronger impact onobjective performance (p<0.001) than subjective performance (p<0.05). As a result,transnationality index explains 43.4% of the variance of objective performance (R 2) andonly 7.2% of subjective performance. Once objective performance is measured mainlyby a ratio of foreign over total (Ebtida index), it is possible that firms that have a largeproportion of its revenues, assets and employees out of borders, also have a greatpercentage of its profit abroad. The positive signs of the other objective performanceitems show that internationalization indeed brings financial returns to companies,though it’s still modest. Perhaps that is why satisfaction with performance does notincrease as much as objective, when firms become more internationalized. Still, thepositive sign and the 5% significance of the relationship between transnationality indexand subjective performance points out to the eminent importance of internationalizationon firm’s results. Model 1 achieved good measures of fit, most of them close to the standardsrecommended in the literature. The goodness of fit (GFI) was 0.86 and the adjusted 8
  9. 9. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia Organizacionalgoodness of fit (AGFI) was 0.76. Both these measures indicate that the data fits well tothe model (Tabachnick and Fidel, 2001). Taking into account the degree of parsimony,the PGFI index was 0.52. This shows that there might be too many parameters to beestimated, considering a small sample. Furthermore, residual based fit indices such asRMR and RMSEA were 0.02 and 0.12 respectively. It is important to note that theRMSEA is not completely adequate in small samples (Hu and Bentler, 1999) and thusshould be interpreted with caution. In an attempt to test alternative measures of internationalization, we added twoitems to the transnationality index construct: number of countries and internationalexperience. The results of model 2 are shown in the next figure.------------------------------------------------Insert Figure 2 about here------------------------------------------------ As figure 2 shows, number of countries is significant as an indicator oftransnationality index, though it has a much lower loading than the UNCATD’s items.On the other hand, international experience does not share the same concept oftransnationality when combined with revenues, assets and employees indexes. Despitebeing not significant, interestingly international experience has a negative sign,indicating that more transnationalized Brazilian companies may also be beginners ininternational markets. Indeed, Sucheta, Herrmann, and Perez (2010) argue that evencompanies that have just started global expansion process may achieve what the authorscall an “early international performance”. In addition, the five-item construct has extremely low reliability, withCronbach’s alpha of 0.02, which invalidates this scale. This result provides evidencethat, in this context, UNCTAD’s transnationality index is a better approach to thedegree of internationalization than the five-item construct tested. Thus, adding otherindicators such as number of countries and international experience does not addinternal consistency or explanatory power to the transnationality index. It is interesting to note that when we add the two other internationalization itemsin model 2, transnationality index showed no significant impact on satisfaction(subjective performance). We may infer that companies in international markets forlong periods and in many countries are not more satisfied with foreign performance thanbeginners, whereas companies that have a great percentage of revenues, assets andemployees abroad tend to be more satisfied. Despite achieving relatively similar measures of fit, model 2 is inferior to model1. The goodness of fit (GFI) was 0.85, the adjusted goodness of fit (AGFI) was 0.77 andthe PGFI was 0.57. However, this model showed poor fitting measures on residualbased indices such as RMR (0.44) and RMSEA (0.10).DISCUSSION This study attempted to test two different measures of internationalization and tofurther assess the effect of the degree of internationalization on both objective andsubjective performance. Our purpose was to contrast both types of performancemeasures, assessing which is more affected by the firms’ internationalization and towhat extent. We also tested the effectiveness of additional measures ofinternationalization in an attempt to compare the three-item UNCTAD index toalternative approaches. 9
  10. 10. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia Organizacional From a survey with 41 Brazilian multinational groups and information of twoyears of international activities, we were able to propose and test a model usingStructural Equation Modeling. Our results show that the more internationalized a firm is, the better it performsoverseas. Firms with a higher degree of internationalization were found to be moresatisfied with foreign sales, sales growth, profits and market share. Additionally,internationalization leads to a higher percentage of foreign profits over total, and anincrease in foreign return on sales. On the other hand, return on assets is notsignificantly impacted by the transnationality index. The three indicators of UNCTAD (foreign revenues/total revenues, foreignassets/total assets and foreign employees/total employees) are more effective inpredicting the degree of internationalization in this context than adding other measuressuch as the number of countries and international experience. International experience isnot a significant indicator of the degree of internationalization. Surprisingly, it has anegative sign, indicating that more transnationalized Brazilian companies may also bebeginners in international markets. This finding is consistent with the sample profile, inwhich the top internationalized company, JBS-Friboi (food industry), established thefirst international subsidiary only five years ago. Thus, international experience does notseem to be detrimental for a company to be considered much internationalized. Furthermore, the proportion of foreign employees over total employees wasfound to have a slightly stronger impact on the transnationality index, than the othermeasures. Although the degree of internationalization impacts objective performance toa higher extant than subjective performance, the positive sign and 5% of significanceshows that firms may be more satisfied with foreign performance as their foreign assets,revenues and employees increases. Thus, firms in initial stages of internationalizationmight have little of its performance accounted by foreign activities and might have lowsatisfaction rates. As the commitment with foreign markets increase, especially in termsof assets, revenues and employees, the percentage of foreign profits tend to increase, ascompared to total profits. This study has important implications to both theory and executives. First, itshows that the UNCTAD transnationality index is more reliable in the context ofBrazilian firms than using additional measures. Second, it provides empirical evidencethat the more revenues, assets and employees a BTNC has across borders, as aproportion of total operations, the better its foreign performance. Thus, firms may seethe internationalization strategy as a way to enhance foreign performance. Finally,internationalization increases objective performance and also executive’s satisfactionwith performance. Yet, the effect of internationalization is stronger on objectivemeasures. Nevertheless, it is important to note several limitations of this study. The mainone regards the small sample size. Although 41 firms can be considered fairlyrepresentative of the Brazilian TNCs, the sample is still small for achieving good fitindexes and explanatory power in the structural equation modeling. Furthermore, wewere only able to collect information about two years of international activities. Oncethis is an ongoing project, we plan to increase our sample and obtain information of atleast three years in order to test the relationships on a longitudinal sample. Also, sincethe study was built based on a survey with Brazilian TNCs, researchers should becautious when generalizing to other countries. We thus suggest that this study beexpanded to other countries as a way to attest the representativeness of the results. 10
  11. 11. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia OrganizacionalREFERENCESAl-Khalifa AK, Peterson SE. 2004. On the relationship between initial motivation, andsatisfaction and performance in joint ventures. European Journal of Marketing, 38(1/2):150-174.Anderson JC, Gerbing DW. 1982. Some methods for respecifying measurement modelsto obtain unidimensional construct measurement. Journal of Marketing Research. 19(4):453-460.Andersson U, Forsgren M, Pedersen T. 2001. Subsidiary performance in multinationalcorporations: the importance of technology embeddedness. International BusinessReview, 10 : 3–23.Anthony G, Paul WB. 2003. Geographic scope and multinational enterpriseperformance. Strategic Management Journal, 24(13): 1289–1306.Anurag S. 1998. Mode of entry and ex-post performance. Strategic ManagementJournal, 19(9): 879–900.Bagozzi RP, Yi Y, Philips LW. 1991. Assessing construct validity in organizationalresearch. Administrative Science Quarterly, 36(3): 421-458.Bagozzi RP, Yi Y. 1993. An Examination of the Psychometric Properties of Measuresof Negative Affect in the PANAS-X Scales. Journal of Personality and SocialPsychology. 65(4): pp. 836-851.Barkema H, Shenkar O, Vermulen F, Bell J. 1997. Working abroad, working withothers: how firms learn to operate international joint ventures. Academy of ManagementJournal, 40(2): 426-442.Bobillo, Alfredo M.; Lopez-Iturriaga, Felix; Tejerina-Gaite, Fernando. 2010. Firmperformance and international diversification: The internal and external competitiveadvantages. International Business Review, 19(6), p. 607–618Bouquet C, Morrison A, Birkinshaw J. 2009. International attention and multinationalenterprise performance. Journal of International Business Studies, 40(1):108-131.Brouthers KD, Brouthers LE, Werner S. 2008. Real options, international entry modechoice and performance. Journal of Management Studies. 45(5): 936-960.Bryman A. 1992. Research Methods and Organization Studies. London: Unwin Hyman.Buch, Claudia M.; Kleinert, Jörn; Lipponer, Alexander; Toubal, Farid. 2005. ForeignDirect Investment. Economic Policy January, pp. 51–110.Cadogan, J.W.; Kuivalainen, O.; Sundqvist, S. 2009. Export Market-Oriented Behaviorand Export Performance: Quadratic and Moderating Effects Under Differing Degrees ofMarket Dynamism and Internationalization. Journal of International Marketing, 17(4):pp. 71–89.Churchill GA Jr. 1979. A paradigm for developing better measures of marketingconstructs. Journal of Marketing Research, 16(1): 64-73.Cicic M, Patterson P, Shoham A. 2002. Antecedents of international performance: aservice firms’ perspective. European Journal of Marketing, 36(9/10): 1103-1118.Collins JM. 1990. A market performance comparison of U.S. firms active in domestic,developed and developing countries. Journal of International Business Studies, 2 : 271-87.Colton DA, Roth MS, Bearden WO. 2010. Drivers of international e-tail performance:the complexities of orientations and resources. Journal of International Marketing.,18(1):1-22. 11
  12. 12. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia OrganizacionalDaniels JD, Bracker J. 1989. Profit performance: Do foreign operations make adifference? Management International Review, 29(1): 45-56.Davidson W H.1984. Administrative orientation and international performance.Journal of International Business Studies, 15(2) 11-23.Dorrenbacher C. 2000. Measuring corporate internationalisation: a review ofmeasurement concepts and their use. Intereconomics 35 (3): 119-26.Dunning JH. 1996. The geographical sources of the competitiveness of firms: someresults of a new survey. Transnational corporations 5 (3): 1-29.Dunning JH. 2000. The eclectic paradigm as an envelope for economic and businesstheories of MNE activity. International Business Review, 9(2): 163–190.Dymsza W. 1984. Global strategic planning: a model and recent developments. Journalof International Business Studies, Autumn Special Issue on Strategic Planning,Autonomy and Control Processes in Multinational Corporations, 15(2): 169-183.Fisch, JH, Oesterle M-J. 2003. Exploring the globalization of German MNCs with thecomplex spread and diversity measure. Schmalenbach Business Review, Vol. 55,January, pp. 2–21.Frenz, M, Girardone, C, Ietto-Gillies, G. 2005. Multinationality Matters in Innovation:The Case of the UK Financial Services. Industry and Innovation, (12)1: 65–92.Geringer JM, Paul WB, Richard CD. 1989. Diversification strategy andinternationalization: Implications for MNE performance. Strategic ManagementJournal, 10: 109-20.Goerzen A, Beamish PW. 2003. The effect of alliance network diversity onmultinational enterprise performance. Strategic Management Journal, 26: 333–354.Grant RM. 1987. Multinationality and performance among British manufacturingcompanies. Journal of International Business Studies, 18(3): 79-89.Hair JF Jr, Anderson RE, Tathem RL, Black WC. 2005. Análise Multivariada de Dados.Porto Alegre: Bookman.Hennart J, Reddy S. 1997. The choice between mergers/acquisitions and joint ventures:the case of japanese investors in the united states. Strategic Management Journal. Vol.18(1): 1-12.Hennart J, Reddy, S. 1997. The choice between mergers/acquisitions and joint ventures:the case of Japanese investors in the United States. Strategic ManagementJournal¸18(1): 1-12.Hu L-T, Bentler PM, Sundaram A, Black J. 1992. The environment and internalorganization of multinational enterprises. Academy of Management Review, 17:729–757.Hu L-T, Bentler PM. 1999. Cutoff criteria for fit indices in covariance structureanalysis: Conventional criteria versus new alternatives. Structural Equation Modeling,6: 1–55.Hsu C-C, Boggs DJ. 2003. Internationalization and Performance: Traditional Measuresand Their Decomposition. Multinational Business Review, 11(3): 23-49.Ietto-Gillies G. 1998. Different conceptual frameworks for the assessment of the degreeof internationalization: an empirical analysis of the various indices for the top 100transnational corporations. Transnational Corporations, 7 (1): 17–39.Isobe T, Makino S, Montgomery DB. 2000. Resource commitment, entry timing, andmarket performance of foreign direct investments in emerging economies: the case ofjapanese international joint ventures in china. Academy of Management Journal,43(3):468-484. 12
  13. 13. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia OrganizacionalJane W, Lu PWB. International diversification and firm Performance: the s-curvehypothesis. Academy of Management Journal, 47(4): 598–609. 2004.Johanson J, Vahlne J-E. 1977. The internationalization process of the firm-A model ofknowledge development and increasing foreign market commitments. Journal ofInternational Business Studies, 8 (Spring/Summer): 23-32.Kline RB. 2005. Principles and practice of structural equation modeling. 2a Ed. NewYork: The Guilford Press.Kogut B, Singh H. 1988. The effect of national culture on the choice of entry mode.Journal of International Business Studies, 19(3): 411–432.Lel U, Miller DP. 2008. International cross-listing, firm performance, and topmanagement turnover: a test of the bonding hypothesis. Journal of Finance, 63(4):1897-1937.Loncan T, Nique WM. 2010. Degree of Internationalization and Performance: Evidencefrom Emerging Brazilian Multinational Firms. GCG Georgetown University –Universia. Vol. 4 No. 1 ISSN: 1988-7116.Lu JW, Beamish PW. 2004. International diversification and firm performance: the s-curve hypothesis. Academy of Management Journal, 47(4): 598-609.Luo X, Bhattacharya CB. 2003. The debate over doing good: corporate socialperformance, strategic marketing levers, and firm-idiosyncratic risk. Journal ofMarketing, 73(6): 198–213.Makino S, Isobe T. 2003. Sources of foreign affiliate performance. Academy ofManagement Proceedings. pp. H1-H6.Marsh HW, Balla JR, Donald RP. 1988. Goodness-of-fit indexes in confirmatory factoranalysis: the effect of sample size. Psychological Bulletin, 103(3): 391-149.Miller SR, Eden L. 2006. Local Density and Foreign Subsidiary Performance. Academyof Management Journal. Vol. 49 Issue 2, p341-355, 15p.Nielsen B. 2007. Determining international strategic alliance performance: Amultidimensional approach.Copenhagen Business School, Center for Strategy andGlobalization, Porcelaenshaven 24, DK-2000 Frederiksberg, DenmarkOliver, R. L. 1997. Satisfaction: A Behavioral Perspective on the Consumer. McGraw-Hill, International Editions.Outreville, J. François. (2008). Foreign Affiliates of the Largest Insurance Groups:Location-Specific Advantages. The Journal of Risk and Insurance, (75)2: 463-491.Pangarkar N, Klein S. 2004. The impact of control on international joint ventureperformance: a contingency approach. Journal of International Marketing, 12(3): 86-107.Park SH, Ungson GR. 1997. The effect of national culture, organizationalcomplementarity, and economic motivation on joint venture dissolution. Academy ofManagement Journal, 40(2): 279-308.Ramaswamy K. 1993. Multinationality and performance: an empirical examination ofthe moderating effect of configuration. Academy of Management Proceedings, 142-146.Ramsey JJ, Barakat LL, Cruz LC, Cretoiu SL. 2010. Transnationality Ranking ofBrazilian Companies: Rethingking Global Strategy. Fundação Dom Cabral, June.Ramsey JJ, Barakat LL. 2009. Transnationality Ranking of Brazilian Companies:Internationalization On The Aftermath Of The Global Financial Crisis. Fundação DomCabral, June. 13
  14. 14. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia OrganizacionalRugman, AM, Oh, CH. 2010. Does the regional nature of multinationals affect themultinationality and performance relationship? International Business Review, 19(5): p.479–488.Ruzzier, M, Antoncic, B, Hisrich, RD. (2007). The internationalization of SMEs:developing and testing a multi-dimensional measure on Slovenian firms.Entrepreneurship & Regional Development, (19)March: p.161–183.Shoham, A. 1998. "Export Performance: A Conceptualization and EmpiricalAssessment." Journal of International Marketing, 6 (3): 59-81.Sousa, CMP Export performance measurement: an evaluation of the empirical researchin the literature. Academy of Marketing Science Review, Volume 2004 no. 09 Available: N, Herrmann P, Perez P. 2010. Domestic mindsets and early internationalperformance: the moderating effect of global industry conditions. StrategicManagement Journal. DOI: 10.1002/smj.888.Sullivan D. 1994. Measuring the degree of internationalisation of a firm. Journal ofInternational Business Studies, 25 (2): 325-342.Tabachnick BG, Fidell LS. 2001. Using multivariate statistics. Allyn & Bacon,Massachusetts.Taggart J, Taggart J.1999. International competitiveness and the single currency.Business Strategy Review, 10(2): 25–36.Tuselmanna, H-J, Allenb, MMC.; Barretta, S, McDonald, F. (2008). Varieties andvariability of employee relations approaches in US subsidiaries: country-of-origineffects and the level and type of industry internationalisation. The International Journalof Human Resource Management, (19)9: 1622–1635.UNCTAD. 1995. World Investment Report 1995. Transnational Corporations andCompetitiveness, Geneva: United Nations.Wall, TD, Michie J, Patterson M, Wood SJ, Sheehan M, Clegg CW, West M. 2004. Onthe validity of subjective measures of company performance. Personnel Psychology,57,95-118.Woodcock CP, Beamish PW, Makino, S. 1994. Ownership-based entry mode strategiesand international performance. Journal of International Business Studies, 25(2): P253-273.Zeria Y, Newburry W, Yeheskel O. 1997. Factors affecting the effectiveness of equityinternational joint ventures (eijvs) in hungary. Management International Review,37(3): 259-279. 14
  15. 15. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia OrganizacionalTable 1- Descriptive statistics Mean s.d 1 2 3 4 5 6 7 8 9 10 11 1. Revenues index 0,20 0,20 2. Assets index 0,15 0,17 0,62** 3. Employees index 0,16 0,18 0,65** 0,74** 4. Ebtida index 0,14 0,15 0,66** 0,38** 0,38** 5. Foreign return on sales (ROS) 0,03 0,19 0,18 0,11 0,18 0,27* 6. Foreign return on assets (ROA) 0,06 0,39 -0,03 0,01 0,02 0,13 0,37** 7. Satisfaction with Sales 3,27 1,04 0,15 0,15 0,06 0,10 0,21 -0,01 8. Satisfaction with sales growth 3,00 1,21 0,13 0,15 -0,03 0,19 0,25* 0,17 0,56** 9. Satisfaction with profits 2,92 0,92 0,04 0,17 0,09 0,16 0,16 0,02 0,46** 0,42** 10. Satisfaction with market share 3,21 0,95 0,26* 0,31* 0,30* 0,24* 0,33* 0,07 0,49** 0,49** 0,38** 11. International experience 18,57 14,05 0,09 -0,06 -0,09 0,03 0,14 -0,07 0,08 0,14 0,02 0,02 12. Number of countries 8,76 7,81 0,14 0,29* 0,24* 0,02 -0,09 -0,09 0,00 0,21 0,02 0,06 0,01Source: Research dataOBS: ***Beta is significant at 0.1% level; **Beta is significant at 1% level; *Beta is significant at 5% level; †Beta is significant at 10% level. 15
  16. 16. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia OrganizacionalFigure 1 – Model 1 – UNCTAD’s transnationality index and objective and subjective performance Ebtida Index 0.78*** Objective 0.38** Foreign Return Performance on Sales R2 =43.4% Revenues Index 0.67*** 0.80*** 0.18 Foreign Return on Assets 0.82*** Transnationality Assets Index Index Satisfaction with 0.74*** Sales Employees 0.84*** Index 0.27* Subjective 0.73*** Satisfaction with Performance Sales Growth R2 =7.2% 0.59*** Satisfaction with Profits 0.68*** Satisfaction with Market Share Source: Research dataOBS: ***Beta is significant at 0.1% level; **Beta is significant at 1% level; *Beta is significant at 5% level; †Beta is significant at 10% level.Figure 2 – Model 2 – Alternative transnationality index and objective and subjective performance 16
  17. 17. Artigo apresentado no XXXV Encontro da ANPAD - 04 a 07 de setembro de 2011Rio de Janeiro/RJ – Indicado a prêmio de melhor artigo da Divisão de Estratégia Organizacional Ebtida Index 0.77*** Countries Objective 0.26* 0.39** Foreign Return Performance on Sales R2 =40.4% International 0.64* Experience -0.03 0.19 Foreign Return on Assets 0.79*** Transnationality Revenues Index Index 0.83*** Satisfaction with 0.74*** Sales Assets Index 0.27 0.85*** Subjective 0.73*** Satisfaction with Performance Sales Growth Employees R2 =7.1% Index 0.59*** Satisfaction with Profits 0.68*** Satisfaction with Market ShareSource: Research dataOBS: ***Beta is significant at 0.1% level; **Beta is significant at 1% level; *Beta is significant at 5% level; †Beta is significant at 10% level. 17